Rapid Data Evacuation Based on Zone Risks for Large-Scale Disasters in Software-Defined Optical Networks
Guozhu Zhao, Lisheng Ma, Shenghui Zhao · 2019
Slos (Service level objectives) defines specific quantization parameters for network service performance such as bandwidth of network, response time, delay, safety requirements, etc. The sum of all the factors that cause a violation of the Slos in a certain zone is called the Zone Risks of the data center network. This paper quantifies the Zone Risks of the data center network, under the SDN network architecture, the RBF Neural Network is used to predict the Zone Risks of the network. According to the level of Zone Risks, the data is placed in data center nodes with the lowest Zone Risks in advance. By this way, the strategy proposed by this paper significantly reduce the total amount of data need to be evacuated and improve the quality and efficiency of data evacuation when the Large-Scale disaster arrives. In order to verify the effectiveness of the proposed strategy, we use Mininet and OpenDaylight as SDN emulation and the controller respectively. The results show that the proposed strategy can dynamically adjust the physical distribution of data before the disaster arrives, reaching the goal of rapid data evacuation.